Conventional systems provide functionality for determining responsiveness of servers to network requests. Typically, static response times are defined by an administrator. If a server fails to respond to a request within the allocated response time, the server may be identified as an unavailable server. Such a system would not typically accommodate variations in response time due to increased load or type of request.
Nor would such a system typically have the flexibility to support different response times needed when responding to requests for different servers or network services on different servers. For example, in a system providing a plurality of network services on a plurality of servers, the system may support only one, static response time while the network services require varying amounts of time to respond to different types of requests. Some servers, or network services on servers, may be available but require additional time than that allotted by the pre-determined, static response time. These servers or network services may be identified as unavailable although they are available.
A system supporting a variable response time and providing accurate identification of the availability of one of a plurality of network services or servers having varying average response times would be desirable